Siffert_fan,
The central, aero-neutral section is already 50cm wide. Plenty wide enough for a driver's footbox, although this is a straw man argument as, again by regulation, the driver's feet have to be fully behind the front axle centreline. This would mean that, as now, they could taper the nosecone well before the wing needed to be attached. The size of the wings would remain unchanged. Take a look at the front wings on, say, the Williams FW14 or McLaren MP4/5, for example. Those aerofoils get a lot closer than 50cm in their midpoints, and the wings are narrower overall as, in period, keeping the active portion of the wing in the (relatively) laminar flow between wheel and bodywork was the preferred solution. These days, because of the mandatory neutral section, designers are forced to use the space available, even though it is less optimal.
I was never disagreeing with you about turbulent flow resulting in lessening of the downforce generated by the front wing, but you were talking about the nosecone lifting on the straight. What the car does on the straight has nothing to do with its cornering power, and the chances of there being enough sustained lifting flow over the nosecone to actually raise the nose of the car sufficiently to make the driver back off is zero. The Mercedes example only required a pulse of upward turbulent flow to coincide with a bump in the track, which then allowed the normal, flat-floor related blowover mechanisms to take over. The noses of the Mercedes cars are also much wider than the 20-30cm F1 nosecones, so the combination of these circumstances make the two very poor analogues.
As for it not being chaotic flow, I think you are mixing up "chaotic" with "random", two very different sets of circumstances. Random implies unpredictable and, essentially, isotropic. Chaotic implies unstable and sensitive to very small changes in the boundary conditions. Given infinite boundaries and highly stable flow dynamics, even a turbulent airflow can develop metastable structure. This isn't the case for a car, however, as they are much too close to the ground and change direction on a millisecond timescale. Hence, the airflow behind the car is very clearly chaotic, with fluctuations and pressure changes governed by the smallest of changes in ride height, exhaust flow, car attitude and so forth. This is why CFD still isn't able to properly model turbulent flow, as the system is so unstable. One millisecond your local flow is upward, the next downward, half a second later away from you, or whatever.
For a little more technical detail on how F1 teams are maximising the diffuser feed using the high-nose solution, take a look at this:
http://mccabism.blogspot.com/2011/12/how-red-bull-create-streamwise.html